Have me in English! Page professionnelle
 Groupe de travail Apprentissage Projet ANR EXPLO/RA Page personnelle
Accueil CV / Thèmes de recherche Publications Enseignement Liens Contact
Gilles Stoltz

Chargé de recherche CNRS à l'Ecole normale supérieure
Professeur affilié HEC Paris
Membre de l'équipe-projet CLASSIC de l'INRIA

Vous pouvez également retrouver (une partie de) cette liste sur HAL.
This list is (partially) available on the French open e-print archive HAL.


Manuscrit d'habilitation / Habilitation thesis

En français -- Contributions à la prévision séquentielle de suites arbitraires : applications à la théorie des jeux répétés et études empiriques des performances de l'agrégation d'experts
English version -- Contributions to the sequential prediction of arbitrary sequences: applications to the theory of repeated games and empirical studies of the performance of the aggregation of experts

Exposé effectué lors de la soutenance / Slides of the presentation (in French)

Manuel de cours / Textbook (in French)

[20] Vincent Rivoirard and Gilles Stoltz, Statistique en action, Vuibert, first edition, 2009; second edition, 2012.
           550 pages, out of which 450 are included in the hard copy and 100 can be freely downloaded. 

Ouvrage édité / Edited volume

[19] Pierre Alquier, Eric Gautier, and Gilles Stoltz, editors, Inverse Problems and High-Dimensional Estimation,
        Stats in the Château summer school (revised lectures). Springer, Lecture Notes in Statistics, 2011.

Articles soumis / Submitted articles

[18] Nicolò Cesa-Bianchi, Pierre Gaillard, Gábor Lugosi, and Gilles Stoltz, A new look at shifting regret, 2012.

[17] Marie Devaine, Pierre Gaillard, Yannig Goude, and Gilles Stoltz, Forecasting the electricity consumption by aggregating specialized experts; a review of the sequential aggregation of specialized experts, with an application to Slovakian and French country-wide one-day-ahead (half-)hourly predictions [shorter and revised version, original longer version], 2012.
Here is the first full technical report on which the paper is based and here is the second such technical report.


Articles publiés dans des conférences internationales / Articles in proceedings of international conferences

Les articles [1], [2], [5] et [6] ont été présentés à COLT avant d'être complétés et de paraître dans des journaux. L'article [10] a été présenté à ALT. L'article [11] a été présenté à NIPS.
Extended abstracts of the articles [1], [2], [5], and [6] were published in the Proceedings of COLT; an extended abstract of [10] was published in the Proceedings of ALT; an extended abstract of [11] was published in the Proceedings of NIPS.

[16] Sébastien Bubeck, Gilles Stoltz, and Jia Yuan Yu, Lipschitz bandits without the Lipschitz constant, ALT, 2011.

[15] Shie Mannor, Vianney Perchet, and Gilles Stoltz, Robust approachability and regret minimization in games with partial monitoring, COLT, 2011.
The link above is to an extended version submitted for journal publication. Original version published in the COLT proceedings.

[14] Odalric-Ambrym Maillard, Rémi Munos, and Gilles Stoltz, A finite-time analysis of multi-armed bandits problems with Kullback-Leibler divergences, COLT, 2011.

[13] Gábor Lugosi, Omiros Papaspiliopoulos, and Gilles Stoltz, Online multi-task learning with hard constraints, COLT, 2009.


Articles publiés dans des journaux / Articles in journals 

[12] Tomasz Michalski and Gilles Stoltz, Do countries falsify economic data strategically? Some evidence that they might, The Review of Economics and Statistics, accepted for publication.

[11] Sébastien Bubeck, Rémi Munos, Gilles Stoltz, and Csaba Szepesvari, X-armed bandits, Journal of Machine Learning Research, 12(May):1655-1695, 2011.
An earlier version appeared in the Proceedings of NIPS'08.

[10] Sébastien Bubeck, Rémi Munos, and Gilles Stoltz, Pure exploration for multi-armed bandit problems, Theoretical Computer Science, 412:1832-1852, 2011.
An earlier version appeared in the Proceedings of ALT'09. 

[9] Gilles Stoltz, Agrégation séquentielle de prédicteurs : méthodologie générale et applications à la prévision de la qualité de l'air et à celle de la consommation électrique, Journal de la Société Française de Statistique, 151(2):66-106, 2010.
Article de survol écrit suite à la réception du prix Marie-Jeanne Laurent-Duhamel.  / Survey paper, in French.

[8] Shie Mannor and Gilles Stoltz, A geometric proof of calibration, Mathematics of Operations Research, 35:721-727, 2010.

[7] Vivien Mallet, Gilles Stoltz, and Boris Mauricette, Ozone ensemble forecast with machine learning algorithms, Journal of Geophysical Research, 114, D05307, doi:10.1029/2008JD009978, 2009.
Here is the full technical report on which the paper is based.  Slides are also available (talk given in Ottawa for the Marie-Jeanne Laurent-Duhamel prize).  

[6] Shie Mannor, Gábor Lugosi, and Gilles Stoltz, Strategies for prediction under imperfect monitoring, Mathematics of Operations Research, 33:513-528, 2008.
An earlier version appeared in the Proceedings of COLT'07. Slides of talks given at COLT'07 and at 'Entente cordiale' workshop.

[5] Nicolò Cesa-Bianchi, Yishay Mansour, and Gilles Stoltz, Improved second-order bounds for prediction with expert advice, Machine Learning, 66:321-352, 2007. 
This is an invited paper: An earlier version won a Student Machine Learning Award at COLT'05. Slides of the talk given at  COLT'05.

[4] Gilles Stoltz and Gábor Lugosi, Learning correlated equilibria in games with compact sets of strategies, Games and Economic Behavior, 59:187-208, 2007.
Slides of a presentation of this paper at the second world congress of the Game Theory Society. In French: Présentation faite lors du séminaire de théorie des jeux à l'IHP.

[3] Nicolò Cesa-Bianchi, Gábor Lugosi, and Gilles Stoltz, Regret minimization under partial monitoring, Mathematics of Operations Research, 31:562-580, 2006.
Slides of a presentation of this paper at the CIRM in  Marseille. In French: Présentation faite pour le groupe de travail Apprentissage à l'ENS.

[2] Nicolò Cesa-Bianchi, Gábor Lugosi, and Gilles Stoltz, Minimizing regret with label-efficient prediction, IEEE Transactions on Information Theory, 51:2152-2162, 2005.
An earlier version appeared in the Proceedings of COLT'04.  No English slides for this article: Présentation faite lors d'un congrès à Aussois.

[1] Gilles Stoltz and Gábor Lugosi, Internal regret in on-line portfolio selection, Machine Learning, 59:125-159, 2005.
This is an invited paper: An earlier version won a Machine Learning Award at COLT'03. Slides of the talk given at  COLT'05. Slides of a presentation of this paper at the CIRM in Marseille. In French: Présentation faite lors d'un exposé de séminaire à Nice.



Rapports techniques, à transformer en articles (un jour) / Technical reports to be written up in research papers (some day)

Sébastien Gerchinovitz, Vivien Mallet, and Gilles Stoltz, A Further Look at Sequential Aggregation Rules for Ozone Ensemble Forecasting, 2008.

Gilles Stoltz, On the van Trees inequality in a Le Cam's framework, 2001.
Shortnote written for David Pollard during the international semester '2001, a statistics odyseey' in Paris. Writting this research work was the examination! So it has already shown extremely useful.
Here is a readable compact version (in French, though). The English version has to be rewritten entirely.  


Thèse de doctorat / PhD thesis

Gilles Stoltz, Incomplete information and internal regret in prediction of individual sequences, PhD dissertation, Université Paris-Sud, 2005.
All chapters but the first one are in English. The pdf file has a sieze of 1,912 Kbytes. Slides of the defense are also available
You even get to see some pictures of the defense (and of the party at night!): with my twin brother, end of my presentation, jury (1), jury (2), party (0), party (1), party (2).